Deep supervised learning algorithms typically require a large volume of labeled data to
achieve satisfactory performance. However, the process of collecting and labeling such data …

In recent years, neural architecture-based recommender systems have achieved
tremendous success, but they still fall short of expectation when dealing with highly sparse …

Despite the success of large vision and language models (VLMs) in many downstream
applications, it is unclear how well they encode compositional information. Here, we create …

Manual annotation of large-scale point cloud dataset for varying tasks such as 3D object
classification, segmentation and detection is often laborious owing to the irregular structure …

We introduce OpenShape, a method for learning multi-modal joint representations of text,
image, and point clouds. We adopt the commonly used multi-modal contrastive learning …

Unsupervised localization and segmentation are long-standing computer vision challenges
that involve decomposing an image into semantically-meaningful segments without any …

Current semantic segmentation methods focus only on mining" local" context, ie,
dependencies between pixels within individual images, by context-aggregation modules …

Self-supervised learning has gained popularity because of its ability to avoid the cost of
annotating large-scale datasets. It is capable of adopting self-defined pseudolabels as …

Heterogeneous graph neural networks (HGNNs) as an emerging technique have shown
superior capacity of dealing with heterogeneous information network (HIN). However, most …

A vital step towards the widespread adoption of neural retrieval models is their resource
efficiency throughout the training, indexing and query workflows. The neural IR community …